nuclear mutants of maize with defects in chloroplast polysome

The Plant Cell, Vol. 5, 389-402, April 1993 O 1993 American Society of Plant Physiologists

Nuclear Mutants of Maize with Defects in Chloroplast Polysome Assembly Have Altered Chloroplast RNA Metabolism

Alice Barkan lnstitute of Molecular Biology, University of Oregon, Eugene, Oregon 97403-1229

The molecular basis for the photosynthetic defect in four nuclear mutants of maize was investigated. Mutants hcf7, cpsl-1, cpsí-2, and cps2 contained reduced levels of many chloroplast-encoded proteins without corresponding deficiencies in chloroplast mRNAs. Many chloroplast mRNAs were associated with abnormally few ribosomes, indicating that the protein deficiencies were due to global defects in chloroplast translation. These mutants were used to study the effects of reduced ribosome association on the metabolism of chloroplast RNAs. The level of the fbcL mRNA was reduced fourfold in each mutant, but was unaltered in other nonphotosynthetic mutants with normal chloroplast translation. These results suggest that the rbcL mRNA is destabilized as a consequence of its decreased association with ribosomes. The fact that many other chloroplast mRNAs accumulated to normal levels demonstrated that a decreased association with ribosomes does not significantly alter their stabilities or processing. hcf7 seedlings had a gross defect in the processing of the 16s rRNA: the primary lesion in this mutant may be a defect in 16s rRNA processing itself.

INTRODUCTION

A vital component of leaf cell differentiation is the differentiation of proplastids into chloroplasts. The proplastid is a spherical organelle with little interna1 membrane structure, few ribosomes, and little pigmentation. During the differentiation of leaf cells, the proplastid differentiates into the much larger, more elongated chloroplast containing the abundant and highly or- ganized thylakoid membrane (reviewed by Mullet, 1988). Accompanying these morphological changes are changes in lipid content (Leech et al., 1973), increases in the abundance of ribosomes and photosynthetic enzymes (reviewed by Mullet, 1988), transient increases in DNA copy number and transcriptional activity (Baumgartner et al., 1989), and an increase in the level of spliced mRNAs relative to their unspliced precursors (Barkan, 1989).

While the events accompanying chloroplast differentiation have been described in considerable detail, little is known about ways in which products of nuclear genes integrate these events with changes in the rest of the cell. Few nuclear genes with defined roles in chloroplast gene expression have been identified. Clones corresponding to a number of nuclear genes encoding chloroplast ribosomal proteins (reviewed by Subramanian et al., 1991) and translation factors (Baldauf et al., 1990) have been isolated. Severa1 nuclear genes have been identified that encode chloroplast RNA binding proteins (Li and Sugiura, 1991; Schuster and Gruissem, 1991; Cook and Walker, 1992; Mieszczak et al., 1992), but the roles of these proteins in the chloroplast remain unknown.

A potentially powerful way to identify nuclear genes that play critical roles in chloroplast gene expression is to isolate nuclear mutations that block chloroplast function. Numerous nuclear mutants of this type exist (reviewed by Miles, 1982; Somerville, 1986; Taylor et al., 1987; Taylor, 1989), but the primary defects in most of these are unknown. This approach has been exploited most thoroughly in studies of Chlamydo- monas (reviewed by Rochaix, 1992). It is likely that nove1 mechanisms will be unmasked by genetic studies in higher plants, considering the substantial differences between plants and algae in chloroplast gene organization and expression.

To identify and isolate nuclear genes that modulate chloroplast gene expression in higher plants, we are isolating transposon-induced mutants of maize with defects in chloroplast translation and mRNA metabolism. This paper concerns four nuclear mutants (hcf7, cpsl-7, cpsl-2, and cps2) that fail to assemble normal chloroplast polysomes. Mutant seedlings are slightly pale green relative to their normal siblings and photosynthetically inactive. To gain insight into the basis for the photosynthetic defects, the mutant phenotypes were analyzed in detail. The level of ribulose-1 ,Bbisphosphate carboxylase/ oxygenase (Rubisco) in the mutant seedlings was 20- to 40-fold lower than normal. cps7 and hcf7 mutants were deficient in many thylakoid membrane proteins as well. Their pigmentation was nonetheless near normal due to the normal accumulation of the light-harvesting complexes. While most chloroplast mRNAs were of normal size and abundance, they were

390 The Plant Cell

associated with very few ribosomes. Therefore, these mutantsare likely to be defective in the initiation step of translation.

Based upon their phenotypes, the cpsl and cps2 mutantscould have lesions in genes encoding ribosomal proteins orinitiation factors. hcf7 was unique in that it had a severe blockin the maturation of the 16S rRNA; thus, the primary defectin hcf7 may be in rRNA processing itself. The defects in poly-some assembly correlated with altered metabolism of at leastone chloroplast mRNA (rbcL). Therefore, changes in the chlo-roplast translation machinery during chloroplast developmentmight be one factor that modulates the stabilities of a subsetof chloroplast mRNAs.

RESULTS

Mutant Isolation and Genetic Analysis

Three of the mutants used in this study were isolated from maizelines harboring active Mutator (Mu) transposons. Because theyproved to be defective in chloroplast protein synthesis, thesewere named cps mutants. The three mutants arose indepen-dently and defined two complementation groups. Thus, theywere named cps7-7, cps1-2, and cps2. All three cps mutationsare unstable (i.e., dark green sectors arise on pale green mutantleaves), indicating that transposon insertions are responsiblefor the mutant phenotypes. These sectors were typically verysmall and were of similar frequency in all of the mutant materialused for this study. The fourth mutant discussed here, hcf7,was isolated from a population of ethylmethane sulfonate-mutagenized material by D. Miles (University of Missouri,Columbia, Missouri). hcf7complemented both cps7 and cps2,indicating that it defines a separate gene. Crosses with B-Atranslocation stocks (Beckett, 1978) revealed that cps2 mapsto the long arm of chromosome 6 and hcf7 and cps? map tothe long arm of chromosome 1.

hcf7, cpsl, and cps2 Mutants Have Reduced Levelsof Photosynthetic Complexes

Protein gel blots were used to quantify levels of representa-tive protein components of each photosynthetic complex.Dilutions of wild-type protein extracts were analyzed on eachgel to aid in quantifying specific proteins in mutant extracts.Examples of these results are shown in Figure 1.

The level of the rbcL gene product (the large subunit ofRubisco) was reduced more than 20-fold in all four mutants(Figure 1A). The Rubisco level was consistently lower in cps7-2than in cps7-7, suggesting that the cps7-7 allele is a "leakier"allele. Levels of each thylakoid membrane complex contain-ing plastid-encoded components were also reduced in hcf7and cps7 mutants (Figure 1B): subunits of photosystem I(psaA/B), photosystem II (psbA), the cytochrome b6f complex

-rbcL-

-rbcL- —„_ ' ~ ^ I

B

11

- - »

-psaA/B-atpA-

- psbA_

"petD_

apA.B =

-LHCP_ iti ...' «•»<IS

Figure 1. Immunoblot Analysis of Chloroplast Proteins.(A) Level of Rubisco in mutant leaves. Equal amounts (or the indicateddilutions of the wild-type sample) of soluble leaf proteins were analyzed.(B) Levels of specific thylakoid membrane proteins in mutant leaves.Equal amounts (or the indicated dilutions of the wild-type sample) ofmembrane-associated leaf proteins were analyzed.The upper panel in each pair shows the results of probing blots withthe indicated monospecific antisera. The lower panel in each pairshows the same blots stained with Ponceau S to visualize the totalpopulation of bound protein. The identities of several of the stainedprotein bands are labeled. The gels shown are examples of many in-dependent analyses of this type. The signals obtained for one or moreproteins sometimes did not decrease in proportion to the amount ofprotein applied to the wild-type (WT) dilution lanes. Such blots werenot used for quantification of those particular proteins. In (B), for ex-ample, differences in the level of the atpA gene product were moreapparent on the stained filters than on the antibody detections. LHCP,light-harvesting chlorophyll a/b binding protein.

(petD), and the ATP synthase (atpA) accumulated to only 5to 10% of wild-type levels. The cps7-2 mutant had the mostsevere loss of thylakoid proteins, as it did of Rubisco. cps2exhibited only a minor reduction in thylakoid membrane pro-teins. Subunits of photosystems I and II, the cytochrome b6fcomplex, and ATP synthase accumulated to 30 to 50% of nor-mal levels (Figure 1B and data not shown).

The psaD, petA,petC, psbC, and psbB gene products weredecreased to a similar extent in each mutant as the thylakoidmembrane proteins shown in Figure 1 (Taylor et al., 1987; data

Chloroplast Polysome Mutants 391

not shown). The fact that hcf7 and cpsl mutants were defi-cient in both membrane and soluble chloroplast-encodedproteins suggested that they might be defective in some generalaspect of chloroplast gene expression. The loss of the nuclear-encoded PSAD and PETC proteins would be an expected out-come of such a global defect in plastid gene expression,because it has generally been found that the stabilities of thecore subunits of each photosynthetic complex decrease dra-matically when the synthesis of one core subunit is blocked(Schmidt and Mishkind, 1983; Spreitzer et al., 1985; Bennounet al., 1986; Takahashi et al., 1991). Although cps2 mutantshad only minor losses of thylakoid proteins, it became appar-ent that they also had a global translation defect, for reasonsthat will be described below.

Most Plastid mRNAs Are of Normal Size andAbundance in hcf7, cpsl, and cps2 Mutants

? C1 H 55 55 f-p. P, >

*•^ w H 3& P- S P-U, Cj ••* O

atpB]rbcL[

Figure 3. RNA Gel Blot Analysis of rbcL and atpB mRNAs.

An equal amount (or the indicated dilutions of the wild-type [WT] sam-ple) of seedling leaf RNA was analyzed in each lane, psbl and pet2are nuclear mutants that lack, specifically, the photosystem II corecomplex and the cytochrome b6f complex, respectively (A. Barkan,unpublished results). The precise ratio of the two rbcL transcripts variesslightly between different preparations of RNA, indicating that theseminor differences are not due to differences in genotype. The filtershown on the left was reprobed with a 16S rRNA-specific probe (seeFigure 5B).

To address the possibility that the global deficiencies inchloroplast-encoded proteins were due to defects in chloroplast

rps!6

(Nj

& & &

psaA/B

mRNA metabolism, the size and abundance of various plastidmRNAs were analyzed on RNA gel blots. Transcripts of thepetA, psaA/B, psbA, rps16, and atpB genes accumulated tonormal levels in these mutants, as shown in Figures 2 and 3.The slight differences between samples in the levels of psaA/B,psbA, and atpB transcripts were less than twofold and werenot reproducible. Transcripts of the petB, petD, psaC, psbC,psbN, psbH, psbE, psbl, atpF, rps18, rp!33, and rps12 geneswere of normal size and abundance as well (data not shown).This failure to detect defects in mRNA metabolism suggestedthat the protein deficiencies resulted from defects at a transla-tional or post-translational step.

One mRNA (rbcL) whose level was reduced in all four ofthese mutants was detected (Figure 3). As is described be-low, this feature is a hallmark of mutants that do not assemblenormal polysomes and is unlikely to be the sole cause of thereduced accumulation of Rubisco.

7 ^S ^ ^0, Q, jjo o -c; 8&

psbA petAFigure 2. RNA Gel Blot Analysis of Chloroplast mRNAs.Two micrograms of seedling leaf RNA was analyzed in each lane. Thegene-specific probes are described in Methods. The filter labeledpsaA/B was hybridized with a probe derived from the psaB gene; thisprobe detects the single polycistronic transcript encoding both psaAand psaB. WT, wild type.

Chloroplast mRNAs Are Associated with AbnormallyFew Ribosomes in hcf7, cpsl, and cps2

To investigate the possibility that a defect in translation led tothe loss of chloroplast proteins, the association of chloroplastmRNAs with ribosomes was examined. A decrease in the rateof translation initiation would result in a decrease in the aver-age number of ribosomes associated with mRNAs. Seedlingleaf lysates were prepared under conditions that maintain in-tact polysomes and size-fractionated in analytical sucrosegradients. Figure 4 shows the distribution of various chloroplastmRNAs in the gradient fractions, assayed by RNA gel blothybridization.

Cytosolic translation appeared to be unaltered by these mu-tations because mutant seedlings were of normal size and

392 The Plant Cell

WT cpsl-l

cpsl-2

cps2

WT hen>edi mentation —

IfctteFigure 4. Association of Chloroplast mRNAs with Polysomes.

Total extracts of the apical half of the second and third leaves of singleseedlings were fractionated on sucrose gradients. The gradients werefractionated into 12 ([A] and [B]) or 10 ([C]) fractions of equal vol-ume. An equal proportion of the RNA purified from each fraction wasanalyzed by RNA gel blot hybridization. Each mutant was analyzedin parallel with a wild-type (WT) sibling. The same results were ob-tained with a second seedling of each genotype (data not shown). 23S*is a breakdown product of the plastid 23S rRNA. The decrease in theratio of rbcL to other transcripts is characteristic of these mutants (seeFigures 2 and 3). WT, wild type.(A) Association of rbcL and atpB mRNAs with polysomes in cpsJ leaves.The distribution of rRNAs in each gradient, visualized by staining thenylon membrane with methylene blue, is shown in the upper panelin each pair. The lower panels show the results of hybridizing the samefilters with a DNA fragment containing both rbcL and atpB sequences.(B) Association of rbcL, atpB, and psaA/B mRNAs with polysomesin cps2 leaves. The distribution of rRNAs in each gradient, visualizedby staining the nylon membrane with methylene blue, is shown in theupper panels. The middle panels show the results of hybridizing thesame filters with a radioactive DNA fragment containing both rbcL and

morphology during their first few weeks of growth. As expected,the distribution of mutant and wild-type cytosolic rRNAs (25Sand 18S) in the sucrose gradients was similar (Figure 4). Thecytosolic rRNAs were distributed in two broad peaks, the up-per peak corresponding to particles smaller than SOS (i.e.,monosomes or smaller), the lower one corresponding to poly-somes. The plastidic rRNAs (23S, 16S, and 23S*) in thewild-type samples were also distributed in two broad peaks.The majority of Chloroplast RNA molecules sedimenting fiveor more fractions from the top of the gradients were associatedwith polysomes, because they were released to smaller parti-cles (i.e., to the top three to four fractions) by treatment withEDTA or by treatment with puromycin in the presence of highsalt (data not shown).

For each mutant, all Chloroplast mRNAs examined showeda substantial shift toward the top of the gradient. For example,the majority of the atpB, rbcL, and psaA/B transcripts in wild-type leaves cosedimented with polysomes (Figure 4). In cps1-1and cps7-2 mutants, however, the majority of the rbcL and atpBmRNA molecules were no longer associated with multipleribosomes (Figure 4A). The defect in cps2 was less severe:there was a dramatic increase in nonpolysomal rbcL, atpB,and psaA/B mRNAs in this mutant, but a substantial fractionof these mRNAs remained associated with multiple ribosomes(Figure 4B).

Whereas the atpB, rbcL, and psaA/B mRNA molecules werealmost entirely polysome associated in wild-type chloroplasts,a significant proportion of the psbA mRNA molecules werenot (Figure 4C). A similar observation was made during studiesof barley chloroplasts (Klein et al., 1988); its significance isunknown. Nonetheless, the hcf7mutation caused a reductionin the proportion of psbA mRNA molecules that were polysomebound (Figure 4C). The number of ribosomes associated withthe rbcL and atpB mRNAs was also decreased in hcf7 seed-lings (Figure 4C).

All Chloroplast mRNAs assayed were associated with ab-normally few ribosomes in these mutants. In addition to thedata shown in Figure 4, the following mRNAs were assayed:rp!16 mRNA in hcf7; the rps4, psbA.psbB, psbC/D, psbE/F, atpA,atpF, psaA/B, and pet/A mRNAs in cpsJ mutants; and the petAand psbA mRNAs in cps2 (data not shown). It was striking thatcps2 mutants failed to assemble many different ChloroplastmRNAs into polysomes of normal size, despite the fact thatthe proteins encoded by some of these mRNAs accumulatedto near normal levels.

atpB sequences. The bottom panels show the results of hybridizinga duplicate set of filters with a probe derived from the psaB gene.(C) Association of psbA, rbcL, and atpB mRNAs with polysomes inr;cr7 leaves. The psbA mRNA and cytosolic rRNAs were visualizedby consecutive hybridizations of the same filters. The rbcL and afpflmRNAs were visualized by hybridizing a duplicate set of filters witha radioactive DNA fragment containing both rbcL and atpB sequences.


These results indicate that the hcf7, cpsl, and cps2 muta-tions each reduced the average number of ribosomesassociated with most or all Chloroplast mRNAs. This is mosteasily explained by a global decrease in the rate of an initia-tion step, rather than an elongation step, of translation. Areduction in the elongation rate would, by itself, result in anincrease in the number of ribosomes associated with mRNAs,as is observed after treatment of chloroplasts with chloram-phenicol (A. Barkan, unpublished results; Klaff and Gruissem,1991), or eukaryotic cells with cycloheximide (Peltz et al., 1992).

It is of interest that the distributions of the two rbcL mRNAsin the gradients were similar (Figure 4). The larger of theseis the primary transcipt; the smaller is a derivative resultingfrom processing at the 5' end (Hanley-Bowdoin et al., 1985;Mullet et al., 1985). Their similar sedimentation in these gra-dients indicates that the 5' processing does not dramaticallyalter the rate at which translation is initiated on this mRNA.

A Defect in Polysome Assembly Is Accompanied bya Fourfold Decrease in the Level of the rbcL mRNA

Because the hcf7, cpsl, and cps2 mutations cause a decreasein the number of ribosomes associated with ChloroplastmRNAs, these mutants may be useful for revealing effects ofribosome association on the metabolism of plastid mRNAs.The fact that most plastid mRNAs were processed normallyand accumulated to normal levels (Figures 2 and 3; data notshown) suggested that the spacing of ribosomes along themRNA does not affect the metabolism of most ChloroplastmRNAs. Figure 3 shows, however, that the metabolism of rbcLtranscripts was altered, in that the level of the two rbcL mRNAswas reduced fourfold in each mutant.

The reduction in the level of rbcL transcripts is not a conse-quence of the photosynthetic deficiency because mostnonphotosynthetic mutants have normal levels of this mRNA(e.g., psbl and pet2 in Figure 3; Barkan et al., 1986). Amongthe 17 mutants for which we have analyzed the associationof plastid mRNAs with polysomes, there was a perfect corre-lation between decreased polysome size and decreased rbcLmRNA accumulation: all 13 mutants in which plastid mRNAswere incorporated normally into polysomes had normal levelsof rbcL mRNA (Barkan et al., 1986; data not shown); the fourmutants (defining three genes) that were defective in polysomeassembly had a fourfold reduction in the level oirbcL mRNAs(Figure 3). Whereas this reduction could be due either to adecrease in their rate of synthesis or to an increase in theirrate of decay, the simplest explanation is that the rbcL mRNAis destabilized as a consequence of decreased associationwith ribosomes.

Chloroplast rRNAs in hcf7, cpsl, and cps2 Mutants:Pleiotropic or Primary Defects in rRNA Processing?

To assess whether the translation defects in these mutants wereassociated with defects in Chloroplast rRNA synthesis or

processing, the structure and abundance of Chloroplast rRNAswere examined. Figure 5 shows the results of RNA gel blotsin which clones derived from the maize Chloroplast rRNA geneswere used as probes. Although a small fraction of the maizeChloroplast 23S rRNA is recovered intact, most of it is recov-ered as two smaller molecules: a 1.8-kb fragment is derivedfrom the 5' portion of the 23S rRNA, and a 1.2-kb fragmentis derived from the 3' portion (Figure 5A; Kossel et al., 1985).Both of the "23S" fragments were unaltered in size and accu-mulated to normal levels in each mutant. Therefore, thetranslation defects are unlikely to be due to altered metabo-lism of the 23S rRNA.

It is interesting that the mature, intact 23S rRNA was unde-tectable in cpsl and hcf? mutants, while larger transcripts,presumably corresponding to pre-23S rRNA, were detectedat normal or increased levels (Figure 5A). Nonphotosyntheticmutants with normal Chloroplast translation were not affectedin this way, although their ratio of pre-23S rRNA to the matureform was often somewhat higher than normal (see crp2 and

23S-5'

23S- 3' 23S-S'

8&

prel6S

16S

165

Figure 5. RNA Gel Blot Analysis of Chloroplast rRNAs.

An equal amount (or the indicated dilutions of the wild-type [WT] sam-ple) of leaf RNA was analyzed in each lane. thaZ and crp2 are nuclearmutants with defects in thylakoid assembly and Chloroplast mRNA pro-cessing, respectively (A. Barkan, unpublished results).(A) Transcripts of the 23S ribosomal DMA. The 23S 3' probe was a4.3-kb Pstl-EcoRI fragment encoding the 3' half of the maize 23S rRNAand some downstream sequences. The 23S 5' probe was a 3-kb Pstlfragment that encodes the 5' half of the maize 23S rRNA and someupstream sequences.(B) Transcripts of the 16S ribosomal DNA. These results were obtainedby reprobing the filter shown in the left panel of Figure 3. The maizeChloroplast Baml3 fragment was used as the hybridization probe.

394 The Plant Cell

tha2 in Figure 5A). The 23S rRNAs detected in cps2 were in-distinguishable from those in nonphotosynthetic mutants withnormal chloroplast translation.

These results indicate that the hcf7 and cps7 mutationscause a decrease in the rate of 23S rRNA maturation and/ora decrease in the stability of the mature form. Despite the factthat pre-23S rRNAs were detected in the mutants, the "bro-ken" molecules of 1.8 and 1.2 kb appeared to have normaltermini, because they comigrated with their counterparts innormal chloroplasts. Therefore, the formation of the "hiddenbreak" near residue 1800 of the 23S rRNA seems to be cou-pled to the processing of the termini.

Analysis of 16S rRNA revealed a clear distinction betweenhcf? and the other mutants. The mature 16S rRNA in cps7 andcps2 mutants accumulated to near normal levels (Figure 5Band data not shown). However, the hcf7 mutation caused afourfold reduction in the level of this molecule (Figure 5B). Theloss of the mature 16S rRNA was accompanied by an increasein the abundance of precursor forms that were undetectablein wild-type plants. Whereas a very small amount of pre-16SrRNA was also detected in the cps mutants, this amount wasno greater than that observed in nonphotosynthetic mutantswith normal chloroplast translation (e.g., crp2). Therefore, aslight decrease in the rate of maturation of the 16S rRNA maybe a consequence of many photosynthetic defects.

To determine whether the pre-16S rRNA in hcf7 chloroplastsfunctions in translation, its association with polysomes wasassessed. Figure 6 shows that most of the pre-16S rRNA wasexcluded from the polysomal fractions, whereas the matureform was not. Therefore, this pre-16S rRNA does not functionefficiently in translation.

The termini of the pre-16S rRNA molecules found in hcf7plants were mapped by S1 nuclease protection, as shown inFigure 7. Wild-type RNA protected a 410-nucleotide fragmentof the "5' probe," consistent with the previous mapping of the5'end of mature 16S rRNA (Strittmatter et al., 1985). RNA fromhcf7 leaves protected an additional fragment of the 5' probeof ~510 nucleotides. The 5' end of the most prominent pre-16S rRNA is therefore ~100 nucleotides upstream of that of

5'-end 3'-end

sedimentation —

1 2 3 4 5 6 7 8 9 10

-pre 16S-16S

<80S polysomes

Figure 6. Sedimentation of Pre-16S rRNA in Polysome Gradients.A lysate of an hcf? leaf was fractionated on a sucrose gradient, as de-scribed in Methods, and analyzed by RNA gel blot hybridization. Themaize chloroplast Bam13 fragment was used as a probe.

Z E-a 5

-

25 "Cti«cj O

*J ,q ^

^*«

17"

1C 25 "C

tlQ w -c; O

1 1 g £ H•- ••

I

^ 5 <f- U G 35 *i *i a

«• ;

'it.

ni-1490-1433

-880

-517•396

S'-probe3'-probe Taql

ValtRNA

D . r16S

lietRNA

prm-1 transcript

16S RNA

pre-16S rRNAs in hc/7

Figure 7. S1 Nuclease Mapping of the Pre-16S rRNA in hcf7 Leaves.Reactions contained total leaf RNA from phenotypically normal sib-lings (WT), diploid mutant plants (hcf/hcf), mutant plants hemizygousfor chromosome 1L (hcf/0; see text), or no maize RNA (tRNA). RNAwas purified from plants grown at either 25 or 17°C, as indicated. Mo-lecular size (MW) markers of 1433,517, and 396 bp were Hinf I fragmentsof pUC18. One nanogram of each double-stranded probe was run inthe lane marked "probes." The 3' probe was 1490 bp; the 5' probe was880 bp. nt, nucleotides.

the mature 16S rRNA. This end maps near to or at the tran-scription start site for the rRNA gene cluster (Strittmatter etal., 1985), which is 117 nucleotides upstream of the mature5' end.

hcf? RNA protected one additional fragment of the 5' probe,which was slightly larger than the fragment protected by ma-ture 16S rRNA. In analyses of the most severely affected plants(hcf/0 grown at 17°C; see below), this band was barely visiblewhile the band corresponding to the fully processed 5' endwas undetectable. This fragment may correspond to a minorprocessed end in wild-type plants that maps 30 nucleotidesupstream of the mature end (Strittmatter et al., 1985).

The maturation of the 3' end of the 16S rRNA was also aber-rant in hcf7. Wild-type RNA protected a 1090-nucleotidefragment of a probe spanning the 3' end of the 16S rRNA (Fig-ure 7). This is consistent with the deduced 3' end of the mature16S rRNA (Schwarz and Ko'ssel, 1980). RNA from hcf7 plantsprotected an additional fragment of ~1400 nucleotides. The3' end that this corresponds to is very near to the start of thetRNAlle coding sequence (see map in Figure 7). A substan-tial amount of the 3' probe was fully protected by RNA frommutant seedlings, indicating that many transcripts continue


an unknown distance past the end of the probe used in theexperiment.

Taken together, these results indicate that the pre-16S rRNAthat accumulates in hcf7 is incompletely processed at bothits 5' and 3' ends. Most of the aberrant molecules begin verynear to the transcription start site and end near the RNaseP cleavage site that defines the beginning of tRNAlle down-stream. These termini are consistent with the mobility of thepre-rRNA in agarose gels. The pre-16S rRNA migrated slightlymore slowly than the 18S rRNA and is, therefore, ~400 nucleo-tides longer than the mature 16S rRNA. The Si-mappingexperiments indicate that this is due to an extension at the3' end of ~300 nucleotides and an extension at the 5' end of~100 nucleotides.

The severity of the defect in 16S rRNA processing in hcf?relative to the other mutants cannot be explained simply asa consequence of a tighter block in translation: cpsl-2 mu-tants had a more severe translation defect (Figure 1), butcontained much less pre-16S rRNA and much more mature16S rRNA than hcf? (Figure 58). Therefore, the primary de-fect in hcf? may be in the processing of the 16S rRNA itself.A defect of this nature would result in a decrease in functional30S ribosomal subunits (Figure 6). The primary defects in cpsland cps2 mutants probably lie elsewhere, because thesemutants accumulated no more pre-16S rRNA than other non-photosynthetic mutants with normal chloroplast translation.

The hcf7 Phenotype Is Sensitive to Gene Dosageand Temperature

While mapping the hcf? mutation to a chromosome arm, it be-came apparent that the mutant phenotype was more severein plants carrying one copy of the mutant allele rather thantwo. Plants heterozygous for the mutation were crossed witha series of plants carrying a reciprocal translocation betweenthe B chromosome and various standard chromosome arms.Because the B chromosome frequently undergoes nondisjunc-tion at the second pollen mitosis, progeny plants hemizygousfor the translocated chromosome arm were obtained at highfrequency (Beckett, 1978).

When plants heterozygous for the hcf? mutation were polli-nated by plants carrying TB-1La (a reciprocal translocationbetween the B chromosome and the long arm of chromosome1), 15 of 100 progeny seedlings were small and pale yellow.An example is shown in Figure 8 (hcflQ). Approximately thesame number were equally small, but fully green (+/0 inFigure 8). The morphology of the small seedlings was char-acteristic of plants hemizygous for the 1L chromosome arm(Beckett, 1978).

It seemed likely that the pale yellow seedlings were plantsin which the Permutation had been "unmasked" (i.e., that hadthe genotype hcf?/0) due to the nondisjunction of the B1L chro-mosome. In support of this, no pigment-deficient plants wereobtained when homozygous wild-type plants were crossed bythe same translocation-bearing plants, and crosses of hcf?l+

plants by stocks carrying B-A translocations involving eightother chromosome arms never gave rise to pigment-deficientprogeny.

To test the hypothesis that the pale yellow progeny repre-sented a more severe expression of the hcf? mutation, theaccumulation of pre-16S rRNAs was examined in S1 protec-tion experiments and by RNA gel blot hybridization, as shownin Figures 7 and 9. The pale yellow plants (designated hcf/0)had an even more dramatic increase than hcf?/hcf? plants inthe ratio of the precursor to the mature form of the 16S rRNA.Several albino mutants were analyzed in parallel to evaluatethe possibility that this defect in rRNA processing was simplya consequence of arrested plastid development. Although thealbino mutants contained reduced levels of 16S rRNA, pre-16S rRNA molecules were only barely detectable (Figure 9).Therefore, the severe defect in 16S rRNA processing is ahallmark of the hcf? mutation, and the pale yellow seedlingsalmost certainly are hemizygous for hcf?. Taken together, theseresults strongly suggest that the hcf? mutation maps to thelong arm of chromosome 1. These results also imply that the

25°C

170C

+/+ hcf/hcf hcf/0 +/0Figure 8. Gene Dosage and Temperature Dependence of Pigmenta-tion in hcf? Seedlings.Plants were germinated and grown in a growth chamber at the indi-cated temperature under a 16-hr day (light intensity of 400 nE rrr2

sec"1), 8-hr night cycle. Plants hemizygous for chromosome 1L havedecreased stature (compare +/+ to +/0), as has been previously ob-served (Beckett, 1978). This decrease in stature is independent of thehcf7 genotype. Plants were harvested after 10 days of growth at 25°Cor 20 days of growth at 17°C.

396 The Plant Cell

25°C 17 °c

I u•q

prel6S16S

9 .o

Figure 9. Gene Dosage and Temperature Dependence of 16S rRNAProcessing in hot? Seedlings.Leaf RNA was purified from plant material grown as described in thelegend to Figure 8. An equal amount (or the indicated dilutions of thewild-type [WT] sample) of total RNA was analyzed by RNA gel blothybridization, using the maize chloroplast Barn'13 fragment as a probe.Lanes labeled Iw, wl, and w11 contained RNA purified from albinoseedlings homozygous for Iwl, w1, or w11. The primary defect in Iw1mutants appears to lie in the carotenoid biosynthetic pathway, becausemutant kernels are white, and chlorophyll accumulates in seedlingsgrown under low-intensity light (Bachmann et al., 1973).

genes encoding chloroplast proteins (reviewed by Taylor, 1989).Experiments using inhibitors of chloroplast transcription andtranslation have led to the model that the transmission of thissignal is dependent upon chloroplast transcription (Rapp andMullet, 1991) and translation (Oelmuller et al., 1986) at criticalstages of development. Because the hcf? defect was apparenteven at the very earliest stages of the proplastid-to-chloroplasttransition (data not shown), this model predicts that the hcf?mutation would have pleiotropic effects on the transcriptionof those nuclear genes that are sensitive to the putative chlo-roplast signal.

Because the level of the mRNA encoding the major light-harvesting chlorophyll a/b binding protein has been shownpreviously to be sensitive to the putative chloroplast signal(Mayfield and Taylor, 1984), its level was compared in leavesof hcf7/hcf7, hcf7IO, and wild-type seedlings, as shown in Fig-ure 10. The level of this mRNA was only slightly decreasedin hcf7IO plants grown at 25°C. However, in plants grown at17°C its level was reduced three- and sixfold in hcf7/hcf7 andhcf7IO leaves, respectively. Therefore, the translation defectin its milder expression does not appear to block the trans-mission of the putative chloroplast signal. The more severetranslation block expressed at low temperatures, however, doesappear to block the signal to some extent.

DISCUSSION

The results of these studies indicate that the cps7, cps2, and/7cf7gene products facilitate the translation of most or all chlo-roplast mRNAs. Mutants defective in these genes had reducedlevels of many chloroplast-encoded proteins without corre-sponding deficiencies in chloroplast mRNAs. Chloroplast

mutation is leaky: deficiencies in pigmentation and in 16S rRNAprocessing are both more severe in plants hemizygous for themutant allele (hcf7IO) than in homozygous hcf7/hcf7 plants.

The severity of the hcf7 phenotype was also increased bygrowth at low temperatures. Mutant seedlings grown at 17°Chad substantially less pigment (Figure 8) and had a lowerproportion of mature 16S rRNA (Figures 7 and 9) than plantsgrown at 25°C. These properties are similar to those of sev-eral "virescent" mutants of maize for which growth at lowtemperatures induces chlorosis and loss of plastid ribosomes(Hopkins and Elfman, 1984). Based upon the variety of pheno-types seen with hcf7, it seems quite likely that some ivory, paleyellow, and virescent mutants are allelic to mutants like thosedescribed here.

Expression of Nuclear Genes Encoding ChloroplastProteins in Chloroplast Translation Mutants

It has been postulated that a "signal" from the chloroplast isnecessary for the normal transcription of a subset of nuclear

17 °c 25 °c 17 °C 25 °C

§ O O

IFigure 10. RNA Gel Blot Analysis of Light-Harvesting Chlorophyll a/bBinding Protein mRNA.Equal amounts of total seedling leaf RNA (or the indicated dilutionsof the wild-type [WT] sample) were applied to each lane. In the panelat right, the rRNAs that bound to the nylon filter were visualized bystaining with methylene blue. At left is shown the results of hybridiz-ing the filter with a probe for the major light-harvesting chlorophyll a/bbinding protein (Mayfield and Taylor, 1984). Plants used for these ex-periments were all harvested between 11 AM and 1 PM to avoidfluctuations in transcript levels due to circadian rhythms.


mRNAs in mutant seedlings were associated with abnormally few ribosomes, suggesting a global reduction in the rate of plastidic translation initiation. The genes defined by these mutations may therefore encode components of the chloroplast translation machinery, enzymes that modify the translation machinery, or factors involved in the expression of that machinery.

Analysis of the mutant phenotypes has provided insights into the primary defect in each mutant and into general fea- tures of chloroplast rRNA metabolism, mRNA metabolism, and translation.

Chloroplast rRNA Metabolism

The hcf7 mutant is unique with regard to its severe defect in the maturation of the 16s rRNA. Because the precursors accumulate at the expense of the mature form, the rate of maturation of 16s rRNA in hcf7chloroplasts appears to be reduced. Without a clone of the hcf7 gene, it is not possible to be certain whether this processing defect is a direct consequence of the hcf7 mutation or a pleiotropic consequence of a defect in ribosome assembly/function. However, the fact that cpsl-2 mutants have a more severe translation blockthan hcf7 but accumulate very little pre-16s rRNA argues that the hcf7 gene product functions directly in the processing of the 16s rRNA. Because the pre-16s rRNA does not function efficiently in translation (Figure 6), a defect in 16s rRNA processing would result in adecrease in the number of functional305 ribosomal subunits and, therefore, a decrease in the translation rate.

The hcf7 defect is not a global defect in rRNA maturation because the “broken” 23s rRNA fragments were of normal size and abundance. An RNase 111-like activity may be involved in the maturation of chloroplast rRNAs (Delp and Kossel, 1991), but the hcf7 phenotype is unlikely to result from a defect in a chloroplast homolog of RNase l l l for two reasons: in bacte- ria, RNase III is involved in generating both 23s and 16s rRNAs, and RNase 111-deficient strains of Escherichia colisyn- thesize 16s rRNA with the mature 5’ end at the normal rate (reviewed by Srivastava and Schlessinger, 1990). Mutants like hcf7should be helpful for defining the enzymes that process chloroplast rRNAs.

Rye leaf cells that differentiate at suitably high temperatures accumulate few chloroplast ribosomes and little chlorophyll (reviewed by Feierabend, 1982). A 16s rRNA precursor of similar size to that found in hcf7 has been observed in heat- bleached rye leaves (Subramanian et al., 1991). Perhaps the rye nuclear genome enodes a temperature-sensitive component of the same enzyme that is defective in hcf7.

Alterations in rRNA metabolism shared by several of these mutants are likely to be pleiotropic consequences of a reduction in translation rate or defective ribosome assembly. For example, hcf7and cpsl mutants failed to accumulate mature intact 23s rRNA. Analysis of other nonphotosynthetic mutants revealed that this property is not simply a consequence of a photosynthetic defect. It is likely to be a hallmark of many types of lesions in the chloroplast translation machinery and will be useful for quickly identifying such mutants in the future.

The final stages of bacterial rRNA processing are coupled with ribosome formation and/or translation (Srivastava and Schlessinger, 1990). The properties of these maize mutants suggest that normal maturation of chloroplast rRNAs also de- pends upon a functional translation apparatus.

The bulk of the chloroplast 23s rRNA in higher plants is recovered as several smaller molecules (reviewed by Delp and Kossel, 1991). It is interesting that the broken “23s” molecules derived from mutant and normal leaves were the same size, even though the unbroken molecules in the mutants accumulated only as pre-23s rRNAs. These results suggest that the break does not occur unless the ends are matured, perhaps because the break occurs only in translating ribosomes.

Chloroplast Translation

Analysis of two Chlamydomonas mutants with defects in chloroplast translation led to the hypothesis that ribosomal protein mRNAs may be preferentially translated under conditions of reduced chloroplast protein synthesis (Liu et al., 1989). One of the mutants used in those studies, cr-7, resembled hcf7 in that it was deficient, specifically, in the small ribosomal subunit (Harris et al., 1974). hcf7 and cpsl might be useful for deter- mining whether this model holds true for higher plant chloroplasts. If mRNAs encoding ribosomal proteins were preferentially translated in these maize mutants, this might be reflected by a relatively unaltered association of ribosomal protein mRNAs with ribosomes. However, this was not found to be the case. Sucrose gradient fractionations revealed that two mRNAs encoding ribosomal proteins (rpl76 and rps4) were also associated with abnormally few ribosomes in hcf7 and cpsl mutants, suggesting that these mRNAs are not preferentially translated. It remains possible that other ribosomal protein mRNAs are preferentially translated in these mutants, or that the differences in translation are not reflected by obvious differences in ribosome association.

It should be noted that many ribosomal proteins do appear to accumulate to normal levels in these mutants, despite the global losses of proteins involved in photosynthesis. Protein gel blots probed with antibodies raised against the spinach 50s or 30s ribosomal subunits revealed similar populations of proteins in extracts of hcfi: cpsl, and wild-type leaves (data not shown). Unfortunately, the identities of individual bands were not known, so it is unclear which, if any, of the proteins detected were encoded by chloroplast genes.

Chloroplast mRNA Metabolism

The stabilities of certain chloroplast mRNAs appear to change during the maturation of spinach leaves (Klaff and Gruissem, 1991) and after illumination of dark-grown barley seedlings (Klein and Mullet, 1990). It is not clear what mediates these changes, although a role for specific mRNA binding proteins has been proposed (Stern and Gruissem, 1987). In both

398 The Plant Cell

prokaryotic and eukaryotic cells, certain mRNAs are stabilized and others are destabilized by their association with translating ribosomes (reviewed by Brawerman, 1989). To address the possibility that developmentally induced changes in chloroplast mRNA stabilities might be mediated, in some cases, by changes in their translation, various chloroplast mRNAs were quantified in these mutants. The level of the rbcL mRNA was reduced fourfold in cpsl, cps2, and hcfi’mutants; in contrast, the rbcL mRNA level was normal in 13 other nonphotosynthetic mutants with unaltered polysome recruitment. Among this latter group were three mutants with 10-fold reductions in the Rubisco protein (A. Barkan, unpublished results), indicating that the absence of Rubisco protein per se does not alter the accumulation of its mRNA.

This correlation between polysome incorporation and rbcL mRNA level seems most likely to be due to a change in the stability of the rbcL mRNA rather than a change in its rate of synthesis. There is ample precedent for this, in that numerous prokaryotic mRNAs are destabilized as a consequence of a reduction in their association with ribosomes (Nilsson et al., 1987; Brawerman, 1989; Yarchuket al., 1991). On the other hand, it is difficult to envision a mechanism linking the transcription of, specifically, the rbcL gene with changes in the recruitment of chloroplast mRNAs into polysomes. If there were a plastidencoded transcription factor specific for rbcL, a reduction in chloroplast translation might lead to a specific decrease in the transcription of the rbcL gene. This is unlikely for two reasons. First, treatment of dark-grown barley seedlings with chloramphenicol actually stimulated light-activated rbcL transcription (Klein, 1991). Second, three maize mutants that appear to have global blocks in chloroplast translation elongation exhibit the global protein losses seen in hcfi’and cpsl mutants, but recruit mRNAs into polysomes and have normal levels of rbcL mRNA (A. Barkan, unpublished results). There- fore, a decrease in the association of chloroplast mRNAs with polysomes rather than a reduction in the rate of translation appears to cause the decrease in the level of the rbcL mRNA.

The conclusion that a decrease in polysome recruitment destabilizes the rbcL mRNA may appear to be at odds with the results of Klaff and Gruissem (1991) who found that lincomycin, an antibiotic that decreases the association of chloroplast mRNAs with polysomes, stabilized the rbcL mRNA in spinach leaves. The mutations described here also decrease polysome association of mRNAs, but result in the destabiliza- tion of the maize rbcL mRNA. This difference may simply reflect a difference in chloroplast mRNA metabolism in maize and spinach. However, both sets of results are consistent with the following common mechanism. Lincomycin inhibits the formation of only the first peptide bond in newly initiated polypeptides. After sufficient time to allow run-off of previously initiated ribosomes, this results in mRNAs that are not devoid of bound ribosomes, but rather have one or two that remain near the start codon (Vazquez, 1979). If there were a particularly labile site on the rbcL mRNA very close to the start codon (as is true of some bacterial mRNAs; reviewed by Brawerman, 1989), lincomycin might increase the stability of this mRNA by stalling

a ribosome over that site. The maize mutations described here would lead to increased exposure of this labile site, because the few ribosomes that do initiate translation probably do not remain at the start codon. Thus, the two sets of results together may indicate that an important determinant for the stability of the rbcL mRNA is a labile site near its start codon.

Despite an extensive analysis of chloroplast transcripts in hcfi’, cpsl, and cps2 mutants, no other significant difference in mRNA size or abundance was observed. Minor differences in the levels of certain transcripts or slight changes in the ratios of transcripts from certain transcription units were observed. However, these differences were also observed in nonphotosynthetic mutants with normal chloroplast translation, indicating that they result from the altered physiology in nonphotosynthetic chloroplasts. Therefore, decreased recruitment of chloroplast mRNAs into polysomes has little effect on the stability or processing of most chloroplast mRNAs.

cps2 Causes a General Decrease in Polysome Assembly but a Specific Loss of Rubisco

The specificity of the protein loss in cps2 is intriguing. All chloroplast mRNAs examined were associated with abnormally few ribosomes, suggesting a global decrease in the translation initiation rate. However, the level of Rubisco was decreased 20-fold while the accumulation of many chloroplast-encoded thylakoid membrane proteins was reduced only twofold. How can one explain this discrepancy? We considered the possibilities that the translation defect was specific to bundle sheath chloroplasts (the primary site of Rubisco synthesis) or that bundle sheath chloroplast differentiation was blocked. However, analysis of polysomes in separated bundle sheath and mesophyll cells revealed that polysome assembly was defective in both bundle sheath and mesophyll chloroplasts (J. Mendel-Hartvig and A. Barkan, unpublished data). Furthermore, malic enzyme, another enzyme specific to bundle sheath chloroplasts, accumulated to normal levels in cps2 (data not shown). A defect in the synthesis of the small subunit of Rubisco is unlikely to be responsible for the loss of the large subunit because rbcS mRNAs accumulate normally (data not shown). In addition, the chloroplast chaperonin-60 was found at normal levels (data not shown), eliminating the possibility that its absence leads to the specific loss of Rubisco. Although the reason for the specific loss of Rubisco in cps2 remains a mystery, the following scenarios are consistent with our observations.

Because cps2 mutants accumulated somewhat more Rubisco than hcfi’and cpsl mutants (Figure l), it appears that, of these, cps2 has the “leakiest” translation defect. Thus, the accumulation of Rubisco seems to be a much more sensitive indicator of decreased polysome recruitment than other prominent chloroplast-encoded proteins. The fact that the rbcL mRNA level was decreased as a consequence of decreased polysome association certainly contributes to this effect by amplifying what might otherwise be small decreases in the translation


rate. But the fourfold reduction in the rbcL mRNA level is not sufficient to account for the 10-fold differential between the accumulation of Rubisco and the accumulation of some thylakoid complexes in cps2 mutants. It may be that the rbcL mRNA has a particularly low affinity for ribosomes, such that a small decrease in the ability of ribosomes to initiate translation results in a disproportionate decrease in the rate of rbcL translation. Alternatively, the disproportionate decrease in the level of Rubisco could be explained if many chloroplast-encoded thylakoid proteins are ordinarily synthesized in excess of the amount that is stabilized by incorporation into photosynthetic complexes, whereas the rbcL gene product is not synthesized in as great an excess.

What Are the Functions of the hcf7, cpsl, and cpsl Genes?

To define the biochemical roles of the genes defined by these mutations, it will be essential to clone them. It is unfortunate in this regard that hcf7 is a chemically induced mutation. On the other hand, the cpsl and cps2 mutants are transposon induced: they arose in maize lines harboring active Mu transposons, and they are unstable, exhibiting small revertant sectors of dark-green tissue on pale green mutant leaves. It should therefore be possible to clone these genes by virtue of their association with a Mu transposon. The phenotypes of the cpsl and cps2 mutants are consistent with the possibility that their mutations lie in nuclear genes encoding previously defined components of the chloroplast translation machinery. Even this apparently trivial possibility would, if confirmed, enhance our understanding of the chloroplast translation machinery. Only approximately one-third of the nuclear genes encoding chloroplast ribosomal proteins have been cloned (reviewed by Subramanian et al., 1991). More importantly, there are a number of plastid-specific ribosomal proteins with no bacterial homologs, whose roles in the chloroplast are not un- derstood (Gantt, 1988; Zhou and Mache, 1989; Johnson et al., 1990). A mutant phenotype associated with mutation of any of these structural genes would be invaluable for assigning function.

It is useful to consider the types of mutants that might arise in future screens for transposon-induced mutants defective in chloroplast translation. Among the mutants with global reductions in chloroplast-encoded proteins and decreased chloroplast polysome size will be mutants with defects in chloroplast rRNA processing (like hcV), and possibly mutants with specific defects in the expression of one or more chloroplast genes encoding ribosomal proteins. This latter class could in- clude mutants with defects in the trans-splicing of the rps72 mRNA and mutants with defects in the transcription or mRNA maturation of any of the other chloroplast-localized ribosomal protein genes. Mutants may also arise that are defective in light- activated (Gamble et al., 1989) or developmentally induced translation, rather than simply in constitutive translation.

Because of these exciting possibilities, we are looking forward to recovering more transposon-induced mutants of this type in the future.

METHODS

Mutant Isolation, Propagation, and Genetic Analysis

hcf7 was recovered as a nonphotosynthetic mutant by D. Miles following an ethylmethane-sulfonate mutagenesis performed by G. Neuffer (University of Missouri, Columbia). A brief account of this mutant was presented previously (Taylor et al., 1987). cpsl-7, cpsl-2, and cps2 were recovered from lines harboring active Mu transposons that were propagated by S. Hake, M. Freeling, and coworkers (University of California at Berkeley). These mutants were first identified by their pale green, seedling lethal phenotype. All four mutations are recessive. They were propagated by outcrossing heterozygotes to inbred lines, followed by self-pollination of heterozygotes to recover homozygous mutant seedlings. The mutations were inherited through the pollen and segregated as single Mendelian traits, indicating that they are localized to single nuclear genes. Except where noted othewise, plants used in these studies were grown in a greenhouse for 10 days, at which time the first two leaves were fully expanded and the third leaf was only par- tially expanded. At this stage, the size and morphology of mutant and normal plants were indistinguishable. All plants were harvested between 11 AM and 2 PM. Within seconds after harvesting, plants were frozen in liquid nitrogen. Frozen tissue was stored at -80°C until use.

Complementation tests were performed by intercrossing plants heterozygous for each mutation. Both crosses made between pairs of plants heterozygous for cpsl-7 and cpsl-2 yielded one-quarter pale green, lethal progeny; these mutant progeny had the global losses of chloroplast proteins seen in the two parents. Therefore, cpsl-7 and cpsl-2 do not complement, and are most likely allelic. Numerous crosses involving one wild-type and one heterozygous plant did not yield any mutant progeny.

All four crosses made between pairs of plants hetsrozygous for hcff and cpsl failed to yield any mutant progeny. Therefore, these mutations complement one another and are almost certainly not allelic. A cross between plants heterozygous for cps2 and cpsl failed to yield mutant progeny, indicating that these mutations also complement one another.

Mutations were mapped to chromosome arms by crosses with B-A translocation stocks. The cps2 mutation was mapped to chromosome 6L: crosses of three cps2/+ plants by TB-6Lc pollen yielded pale green lethal seedlings lacking Rubisco. The cpsl mutations were mapped to chromosome 1L: crosses of five cpsl/+ plants by TB-1La pollen yielded pale green seedlings with hypoploid morphology and lacking Rubisco. The hcf7 mutation was mapped to chromosome lL, as described in this paper.

Electrophoretic Fractionation and lmmunological Detection of Protein

Approximately 0.2 g of leaf tissue in 0.5 mL of ice cold homogenization buffer (40 mM P-mercaptoethanol, 100/0 sucrose, 100 mM Tris-HCI, pH 7.2,5 mM EDTA, 5 mM EGTA, 2 WglmL aprotinin, 2 mM phenylmethyl- sulfonyl fluoride) was ground to a fine slurry with a mortar and pestle. Fibrous material was removed by filtration through a glass wool plug

400 The Plant Cell

in a small syringe. Membranes were pelleted by centrifugation at 4OC for 3 min at 12,000 rpm in a microcentrifuge. The supernatant was removed and used as the “solub1e”fraction. Membranes were then rinsed once in homogenization buffer and resuspended to equal chlorophyll concentrations in homogenization buffer.

Protein samples were solubilized by the addition of an equal vol- ume of 7 M urea, 5.8% SDS, 12.5% 0-mercaptoethanol, 25% glycerol, 0.17 M Tris-HCI, pH 6.8, with incubation at 65OC for 15 min. Samples were then electrophoresed on 13% polyacrylamide gels containing 4 M urea, prepared using the buffer system of Laemmli (Laemmli, 1970). Protein was electrophoretically transferred to nitrocellulose (Sambrook et al., 1989) and visualized on the membrane by staining with Ponceau S (Sigma). Antibodylantigen complexes were detected with the ECL system (Amersham International).

In some experiments, the signal obtained with a particular antibody did not decrease in proportion to the decreasing amounts of protein applied to the wild-type dilution lanes. Such blots were not used to quantify that particular protein. At least five seedlings with each mutant genotype were analyzed, and severa1 sets of gels were run for each mutant. Single examples are shown, due to space limitations.

RNA Gel Blot Analysis

Total leaf RNA was purified from 10-day-old seedlings as described by Martienssen et al. (1989). RNA was fractionated in agarose- formaldehyde gels as described previously (Barkan, 1988), transferred to nylon membrane by capillary blotting, and fixed to the membrane by UV cross-linking. The rRNA bound to membranes was stained with methylene blue by immersing the membrane briefly in a solution of 0.03% methylene blue, 0.3 M NaAc, pH 5.2, and destained by rinsing in water. Hybridizations took place at 65°C in 7% SDS, 0.5 M sodium phosphate, pH 7, and 1 mM EDTA. Membranes were washed for 1.5 hr at 65OC in five changes of 0.2 x SSC (1 x SSC is 0.15 M NaCI, 0.015 M sodium citrate), 0.5% SDS (for DNA probes), or 0.1 x SSC, 0.5% SDS (for RNA probes).

Transcripts of the psaNB, pefA, pet6, pefD, psbC, psbH, psbE, and afpF genes were detected by hybridization with radiolabeled, gel- purified DNA fragments that mapped interna1 to the coding region of the indicated maize chloroplast gene. The specific DNA fragments used have been described previously (Barkan et al., 1986; Barkan, 1988, 1989). TherbcL and afp6/€ transcripts were detected by hybridization with the maize chloroplast Bam9 fragment. DNA probes were radiolabeled by random hexamer priming.

RNA probes were used to detect transcripts of the psbA, rps76, psaC, psbN, psbl, rps78, rpl33, and rps72 genes. The following DNA fragments were used as templates for the production of RNA probes: psbA, a 1.2-kb Bglll-Xbal fragment of spinach chloroplast DNAthat includes most of thepsbA coding sequence; rps76, a 1.4-kb BamHI-EcoRI fragment of the maize chloroplast Bamll fragment; psaC, a 700-bp EcoRl fragment of maize chloroplast DNA that also encodes the N terminus of the ndhD gene; rps72(33, a 1-kb BamHI- Pstl fragment of the maize chloroplast Bam23 fragment; rps72(5’), a 400-bp Hindlll-Kpnl fragment of the maize chloroplast Bam!O fragment; psbN, an 800-bp fragment of the maize chloroplast Bamg’fragment extending between the 3‘end of psb6 and the 5’ end of psbH; psbl, a 200-bp BamHI-EcoRI fragment of the maize chloroplast Bam2 fragment; rps78 and rp/33, a 2-kb Kpnl fragment of the maize chloroplast BamlO fragment. The maize chloroplast BamHl fragments are named as described by Larrinua et al. (1983). RNA probes were transcribed with the appropriate RNA poly- merase in the presence of 32P-UTR as described by the manufacturer.

Polysome Analysis

Polysomes were isolated by grinding -0.3 g of frozen leaf tissue in liquid nitrogen to a fine powder with a mortar and pestle. One milliliter of polysome extraction buffer (0.2 M Tris-HCI, pH 9, 0.2 M KCI, 35 mM MgCI2, 25 mM EGTA, 0.2 M sucrose, 1% Triton X-100, 2% polyoxyethylene-10-tridecyl ether, 0.5 mglmL heparin, 100 mM P-mercaptoethanol, 100 pglmL chloramphenicol, 25 pglmL cycloheximide) was added, and the leaf material was ground further until thawed. After 10 min on ice, nuclei and insoluble material were pelleted by centrifugation for 5 min at 12,000 rpm in a microcentrifuge. Sodium deoxycholate was added to the supernatant to a concentra- tion of 0.5%, and the sample was placed on ice for 5 min. Remaining insoluble material was then pelleted by centrifugation at 12,000 rpm for 15 min. The supernatant (0.4 mL) was layered onto 4.4-mL sucrose gradients that were prepared, centrifuged, and fractionated as described previously (Barkan, 1988). RNA was purified from each fraction by de- tergent treatment, phenol extraction, and ethanol precipitation (Barkan, 1988). Each mutant was analyzed in parallel with a normal sibling. Samples were kept at 4OC throughout preparation.

S i Nuclease Mapping

The probe for the 5‘ end of the pre-16s rRNA was an 880-bp Hincll fragment including the rRNA promoter and extending ~ 5 0 0 bp into sequences encoding the mature 16s rRNA (Schwarz and Kossel, 1980; Strittmatter et al., 1985). The 3’ probe was a 1490-bp fragment with one end at the Hincll site in the middle of the 16s coding region and the other end at a Taql site in the frnl intron (Koch et al., 1981). This DNAfragment extends ~ 5 0 0 bp past the 3’end of the mature 16s rRNA.

Twenty nanograms of gel-purified probe fragment, 100 ng (hcV/hcf7 and hcW0 at 25O) or 50 ng (remaining samples) of total leaf RNA, and 6 pg of yeast tRNA were suspended in 15 pL of SI hybridization buffer (80% formamide, 400 mM NaCI, 40 mM Pipes, pH 6.4, 1 mM EDTA). The mixture was heated to 80°C for 5 minto denature the DNA, cooled to 58OC, and incubated overnight. These conditions prevented rean- nealing of the DNA probe, but allowed the formation of RNAlDNA hybrids. S1 reaction buffer (150 pL; 250 mM NaCI, 30 mM NaAc, pH 5.5, 1 mM ZnSO.,, 10 pglmL denatured calf thymus DNA) containing 150 units of S1 nuclease (Bethesda Research Laboratories) was added to each hybridization and incubated at 37% for 1 hr. Reactions were terminated by the addition of 40 pL of 2.5 M NH,Ac, 50 mM EDTA, 400 pglmL tRNA. Nucleic acids were precipitated by the addition of 2 volumes of ethanol and resuspended in 15 pL of 10 mM Tris, pH 8, 1 mM EDTA. Samples were electrophoresed on 1.5% agarose gels, denatured in alkali, and transferred to nylon membranes. Blots were hybridized with the Baml3fragment of maize chloroplast DNA, which encodes all of the mature 16s rRNA.

ACKNOWLEDGMENTS

I am grateful to Don Miles for providing hcf7 seed and to Sarah Hake, Michael Freeling, and coworkers for access to their Mu stocks. Anti- bodies were generously provided by Bill Taylor (rbcL, psaA/6, and atpA), Gadi Schuster WsbA), Tim Nelson (malic enzyme), Tony Gatenby (GroEL), and Regis Mache (spinach ribosomal proteins). Technical as- sistance was provided by Kirsten Munckand Macie Walker. Iam grateful


to Vicki Chandler, Rodger Voelker, Rob Martienssen, and David Johnson for helpful comments on the manuscript. This work was supported by Grant No. DE-FG06-9lER20054 from the Department of Energy.

Received January 13, 1993; accepted March 1, 1993.

REFERENCES

Bachmann, M., Robertson, D., Bowen, C., and Anderson, 1. (1973). Chloroplast ultrastructure in pigment-deficient mutants of Zea mays under reduced light. J. Ultrastruct. Res. 45, 384-406.

Baldauf, S.L., Manhart, J.R., and Palmer, J.D. (1990). Evolutionary transfer of the chloroplast tufA gene to the nucleus. Nature 344,

Barkan, A. (1988). Proteins encoded by a complex chloroplast transcription unit are each translated from both monocistronic and polycistronic mRNAs. EMBO J. 7, 2637-2644.

Barkan, A. (1989). Tissue-dependent plastid RNA splicing in maize: Transcripts from four plastid genes are predominantly unspliced in leaf meristems and roots. Plant Cell 1, 437-445.

Barkan, A,, Miles, D., and Taylor, W. (1986). Chloroplast gene expression in nuclear, photosynthetic mutants of maize. EMBO J. 5,

Baumgartner, B.J., Rapp, J.C., and Mullet, J.E. (1989). Plastid transcription activity and DNA copy number increase early in barley chloroplast development. Plant Physiol. 89, 1011-1018.

Beckett, J.B. (1978). BA translocations in maize. J. Hered. 69,27-36. Bennoun, P., Spierer-Herz, M., Erickson, J., Girard-Bascou, J.,

Pierre, Y., Delosme, M., and Rochaix, J.-D. (1986). Characteriza- tion of photosystem II mutants of Chlamydomonas reinhardtii lacking the psbA gene. Plant MOI. Biol. 6, 151-160.

Brawerman, G. (1989). mRNA decay: Finding the right targets. Cell 57, 9-10,

Cook, W.B., and Walker, J.C. (1992). ldentification of a maize nucleic acid-binding protein (NBP) belonging to a family of nuclear-encoded chloroplast proteins. Nucl. Acids Res. 20, 359-364.

Delp, G., and Kossel, H. (1991). rRNAs and rRNA genes of plastids. In The Molecular Biology of Plastids, L. Bogorad and I.K. Vasil, eds (San Diego: Academic Press), pp. 139-160.

Feierabend, J. (1982). lnhibition of chloroplast ribosome formation by heat in higher plants. In Methods in Chloroplast Molecular Biology, M. Edelman, R.B. Hallick, and N.-H. Chua, eds(NewYork: Elsevier),

Gemble, P.E., Klein, R.R., and Mullet, J.E. (1989). lllumination of eight- day-old dark-grown barley seedlings activates chloroplast protein synthesis; evidence for regulation of translation initiation. In Pho- tosynthesis, W.R. Briggs, ed (New York: Alan R. Liss), pp. 285-298.

Gantt, J.S. (1988). Nucleotide sequences of cDNAs encoding four com- plete nuclear-encoded plastid ribosomal proteins. Curr. Genet. 14,

Hanley-Bwdoin, L., Orozco, E.M., and Chua, N.-H: (1985). In vitro synthesis and processing of a maize chloroplast transcript encoded by the ribulose I,5-bisphosphate carboxylase large subunit gene. MOI. Cell. Biol. 5, 2733-2745.

262-265.

1421-1427.

pp. 671-680.

519-528.

Harris, E.H., Boynton, J.E., and Gillham, N.W. (1974). Chloroplast ribosome biogenesis in Chlamydomonas. J. Cell Biol. 63, 160-179.

Hopkins, W.G., and Elfman, B. (1984). Temperature-induced chloroplast ribosome deficiency in virescent maize. J. Hered. 75,207-211.

Johnson, C.H., Kruft, V., and Subramanian, A.R. (1990). Identifica- tion of a plastid-specific ribosomal protein in the 30s subunit of chloroplast ribosomes and isolation of the cDNA clone encoding its cytoplasmic precursor. J. Biol. Chem. 265, 12790-12795.

Klaff, P., and Gruissem, W. (1991). Changes in chloroplast mRNA stability during leaf development. Plant Cell 3, 517-529.

Klein, R.R. (1991). Regulation of light-induced chloroplast transcription and translation in eight-day-old dark-grown barley seedlings. Plant Physiol. 97, 335-342.

Klein, R., and Mullet, J. (1990). Light-induced transcription of chloroplast genes. J. Biol. Chem. 265, 1895-1902.

Klein, R.R., Mason, H.S., and Mullet, J.E. (1988). Light-regulated translation of chloroplast proteins. I. Transcripts of psaA-psaB,psbA, and rbcL are associated with polysomes in dark-grown and illumi- nated barley seedlings. J. Cell Biol. 106, 289-301.

Koch, W., Edwards, K., and Kossel, H. (1981). Sequencing of the 16s-23s spacer in a ribosomal RNA operon of Zea mays chloroplast DNA reveals two split tRNA genes. Cell 25, 203-213.

Kijssel, H., Natt, E., Strittmatter, G., Fritzsche, E., Gozdicka- Jozefiak, A., and Przybyl, D. (1985). Structure and expression of rRNA operons from plastids of higher plants. In Molecular Form and Function of the Plant Genome, L. van Vloten-Doting, G. Groot, and T. Hall, eds (New York: Plenum), pp. 183-198.

Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.

Larrinua, I.M., Muskavitch, K.M.T., Gubbins, E.J., and Bogorad, L. (1983). A detailed restriction endonuclease site map of the Zea mays plastid genome. Plant MOI. Biol. 2, 129-140.

Leech, R.M., Rumsby, M.G., and Thompson, W.W. (1973). Plastid differentiation, acyl lipid, and fatty acid changes in developing green maize leaves. Plant Physiol. 52, 240-245.

Li, Y., and Sugiura, M. (1991). Three distinct ribonucleoproteins from tobacco chloroplasts: Each contains a unique amino terminal acidic domain and two ribonucleoprotein consensus motifs. EMBO J. 9,

Liu, X.-Q., Hosler, J.P., Boynton, J.E., and Gillham, N.W. (1989). mRNAs for two ribosomal proteins are preferentially translated in the chloroplast of Chlamydomonas reinhardtii under conditions of reduced protein synthesis. Plant MOI. Biol. 12, 385-394.

Martienssen, R.A., Barkan, A., Freeling, M., and Taylor, W.C. (1989). Molecular cloning of a maize gene involved in photosynthetic membrane organization that is regulated by Robertson’s Motato,: EMBO

Mayfield, S.P., and Taylor, W.C. (1984). Carotenoid-deficient maize seedlings fail to accumulate light-harvesting chlorophyll a/b binding protein (LHCP) mRNA. Eur. J. Biochem. 144, 79-84.

Mieszczak, M., Klahre, U., Levy, J., Goodall, G.J., and Filipowica, W. (1992). Multiple plant RNA binding proteins identified by PCR: Expression of cDNAs encoding RNA binding proteins targeted to chloroplasts in Nicotiana plumbaginifolia. MOI. Gen. Genet. 234,

Miles, D. (1982). The use of mutations to probe phdosynthesis in higher plants. In Methods in Chloroplast Molecular Biology, M. Edelman, R.B. Hallick, and N.-H. Chua, eds (New York Elsevier), pp. 75-106.

3059-3066.

J. 8, 1633-1639.

390-400.

402 The Plant Cell

Mullet, J. (1988). Chloroplast development and gene expression. Annu. Rev. Plant Physiol. Plant MOI. Biol. 39, 475-502.

Mullet, J., Orozco, E., and Chua, N.-H. (1985). Multiple transcripts for higher plant rbcL and atpB genes and localization of the transcription initiation sites of the rbcL gene. Plant MOI. Biol. 4, 39-54.

Nllsson, G., Belasco, J.G., Cohen, S.N., and von Gabain, A. (1987). Effect of premature termination of translation on mRNA stability de- pends on the site of ribosome release. Proc. Natl. Acad. sci. USA

Oelmuller, R., Levltan, I., Bergfeld, R., Rajasekhar, V.K., and Mohr, H. (1986). Expression of nuclear genes as affected by treatments acting on the plastids. Planta 168, 482-492.

Peltr, S., Donahue, J., and Jacobson, A. (1992). A mutation in the tRNA nucleotidyltransferase gene promotes stabilization of mRNAs in Saccharomyces cerevisiae. MOI. Cell. Biol. 12, 5778-5784.

Rapp, J.C., and Mullet, J.E. (1991). Chloroplast transcription is re- quired to express the nuclear genes rbcS and cab. Plastid DNA copy number is regulated independently. Plant MOI. Biol. 17, 813-823.

Rochaix, J.-D. (1992). Post-transcriptional steps in the expression of chloroplast genes. Annu. Rev. Cell Biol. 8, 1-28.

Sambrook, J., Fritsch, E.F., and Maniatls, T. (1989). Molecular Clon- ing: A Laboratory Manual (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press).

Schmidt, G., and Mishkind, M. (1983). Rapid degradation of unas- sembled ribulose 1,5-bisphosphate carboxylase small subunits in chloroplasts. Proc. Natl. Acad. Sci. USA 80, 2632-2636.

Schuster, G., and Gruissem, W. (1991). Chloroplast mRNA 3’ end processing requires a nuclear-encoded RNA-binding protein. EM60

Schwan, Z., and Ktissel, H. (1980). The primary structure of 16s rDNA from Zea mays chloroplast is homologous to E. coli 16s rRNA. Na- ture 283, 739-742.

Somewllle, C. (1986). Analysis of photosynthesis with mutants of higher plants and algae. Annu. Rev. Plant Physiol. 37, 467-507.

Spreltzer, R.J., Goldschmidt-Clermont, M., Rahire, M., and Rochaix, J.-D. (1985). Nonsense mutations in the Chlamydomonas chloroplast

84, 4890-4894.

J. 10, 1493-1502.

gene that codes for the large subunit of ribulose bisphosphate car- boxylaseloxygenase. Proc. Natl. Acad. Sci. USA 82, 5460-5464.

Srivastava, A.K., and Schlessinger, D. (1990). Mechanism and regulation of bacterial ribosomal RNA processing. Annu. Rev. Microbiol.

Stern, D.B., and Gruissem, W. (1987). Control of plastid gene expression: 3’ inverted repeats act as mRNA processing and stabilizing elements, but do not terminate transcription. Cell 51, 1145-1157.

Strittmatter, G., Gozdrickadozefiak, A., and Kossel, H. (1985). Iden- tification of an rRNA operon promoter from Zea mays chloroplasts which excludes the proximal tRNA-Val from the primary transcript.

Subramanian, AR., Stahl, D., and Prombona, A. (1991). Ribosomal proteins, ribosomes, and translation in plastids. In The Molecular Biology of Plastids, L. Bogorad and I.K. Vasil, eds (San Diego: Aca- demic Press), pp. 191-215.

Takahashi, Y., Goldschmidt-Clermont, M., Soen, S.Y., Franzen, L.G., and Rochaix, J.-D. (1991). Directed chloroplast transforma- tion in Chlamydomonas reinhardtii: lnsertional inactivation of the psaC gene encoding the iron sulfur protein destabilizes photosystem I.

Taylor, W. (1989). Regulatory interactions between nuclear and plastid genomes. Annu. Rev. Plant Physiol. Plant MOI. Biol. 40, 211-233.

Taylor, W.C., Barkan, A., and Martienssen, R.A. (1987). Use of nuclear mutants in the analysis of chloroplast development. Dev. Genet.

Vazquez, D. (1979). lnhibitors of Protein Biosynthesis (Berlin: Springer-Verlag).

Yarchuk, O., lost, I., and Dreyfus, M. (1991). The relation between translation and mRNA degradation in the lacZgene. Biochimie 73,

Zhou, D.X., and Mache, R. (1989). Presence in the stroma of chloroplasts of a large pool of a ribosomal protein not structurally related to any fscherichia coli ribosomal protein. MOI. Gen. Genet. 219,

44, 105-129.

EMBO J. 4, 599-604.

EMBO J. 10, 2033-2040.

8, 305-320.

1533-1541.

204-208,

DOI 10.1105/tpc.5.4.389 1993;5;389-402Plant Cell

A. BarkanChloroplast RNA Metabolism.

Nuclear Mutants of Maize with Defects in Chloroplast Polysome Assembly Have Altered

This information is current as of March 27, 2019

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

http://www.plantcell.org/cgi/alerts/ctmain

http://www.plantcell.org/cgi/alerts/ctmain

http://www.aspb.org/publications/subscriptions.cfm

nuclear mutants of maize with defects in chloroplast polysome

Documents